Libgdx - difference between Box2DDebugRenderer.render and spritebatch.draw

Question

I am trying to create a simple Android game using libGDX and physics.

As far as I can understand, we can "draw" physics body objects (box2d body objects) either by:

debugRenderer.render(world, camera.combined); 

or by something like:

  batch.begin();
  sprite = (Sprite) circleBody.getUserData();
  sprite.setPosition(circleBody.getPosition().x - (sprite.getWidth() /2) ,     circleBody.getPosition().y - (sprite.getHeight()/2));
  sprite.setSize(45*2, 45*2);
  sprite.draw(batch);
  batch.end();

Can someone explain me the difference between the two? I think its not ideal to use debugRenderer.render(...) for release/production codes. Is it true?

Answer 1

The difference is very simple. The Box2dDebugRenderer is extremely simple to use. It basically requires just a single line of code (the one you already wrote in your question), and it will render simple shapes to visualize your Box2D World. It will render coloured squares, circles, lines and points for Fixtures, Joints or Contacts.

The second approach you posted is completely selfmade. You can draw whatever you want for the Body. You can draw a simple shape yourself, or a sprite, or maybe an animation or even a 3D model. Maybe even nothing.

That's the "problem" with Box2D. The World you create is purely a physics world and Box2D does not know how to render anything, that's what you have to do yourself. A squared Fixture attached to a Body could be a wall, it could be a box or a door.

This other question on SO has a few pictures which show the difference when both rendering methods are used. (Unfortunately they are rendered on different places, which is what this quesion was about).